{"id":"https://openalex.org/W3195633477","doi":null,"title":"Fastformer: Additive Attention is All You Need","display_name":"Fastformer: Additive Attention is All You Need","publication_year":2021,"publication_date":"2021-08-20","ids":{"openalex":"https://openalex.org/W3195633477","mag":"3195633477"},"language":"en","primary_location":{"id":"mag:3195633477","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2108.09084","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2108.09084","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5001967239","display_name":"Chuhan Wu","orcid":"https://orcid.org/0000-0001-5730-8792","observed_orcids":["https://orcid.org/0000-0001-5730-8792"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chuhan Wu","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0001-5730-8792","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076423724","display_name":"Fangzhao Wu","orcid":"https://orcid.org/0000-0001-9138-1272","observed_orcids":["https://orcid.org/0000-0001-9138-1272"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fangzhao Wu","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0001-9138-1272","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5035067940","display_name":"Tao Qi","orcid":"https://orcid.org/0000-0002-1250-3217","observed_orcids":["https://orcid.org/0000-0002-1250-3217"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tao Qi","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0002-1250-3217","affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100768896","display_name":"Yongfeng Huang","orcid":"https://orcid.org/0000-0003-3825-2230","observed_orcids":["https://orcid.org/0000-0003-3825-2230"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yongfeng Huang","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0003-3825-2230","affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":3,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9286999702453613,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9286999702453613,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.04349999874830246,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.011099999770522118,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/additive-attention","display_name":"additive attention","score":0.9760000109672546},{"id":"https://openalex.org/keywords/efficient-transformers","display_name":"efficient transformers","score":0.796999990940094},{"id":"https://openalex.org/keywords/linear-complexity","display_name":"linear complexity","score":0.7319999933242798},{"id":"https://openalex.org/keywords/global-context-modeling","display_name":"global context modeling","score":0.9409999847412109},{"id":"https://openalex.org/keywords/quadratic-complexity","display_name":"quadratic complexity","score":0.6769999861717224},{"id":"https://openalex.org/keywords/text-understanding","display_name":"text understanding","score":0.5059999823570251}],"concepts":[{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.7370433807373047},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7157272100448608},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7111937403678894},{"id":"https://openalex.org/C129844170","wikidata":"https://www.wikidata.org/wiki/Q41299","display_name":"Quadratic equation","level":2,"score":0.546170175075531},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4186117351055145},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.41433870792388916},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.35981130599975586},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.16499364376068115},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.13002383708953857},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.0777231752872467},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.06065812706947327},{"id":"https://openalex.org/C199539241","wikidata":"https://www.wikidata.org/wiki/Q7748","display_name":"Law","level":1,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C94625758","wikidata":"https://www.wikidata.org/wiki/Q7163","display_name":"Politics","level":2,"score":0.0},{"id":"https://openalex.org/C17744445","wikidata":"https://www.wikidata.org/wiki/Q36442","display_name":"Political science","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:3195633477","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2108.09084","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null}],"best_oa_location":{"id":"mag:3195633477","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2108.09084","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":20,"referenced_works":["https://openalex.org/W1544827683","https://openalex.org/W2142972908","https://openalex.org/W2250539671","https://openalex.org/W2413794162","https://openalex.org/W2626778328","https://openalex.org/W2798240283","https://openalex.org/W2963341956","https://openalex.org/W2964121744","https://openalex.org/W2970793364","https://openalex.org/W2975059944","https://openalex.org/W3015468748","https://openalex.org/W3034236656","https://openalex.org/W3034503922","https://openalex.org/W3034573343","https://openalex.org/W3045733172","https://openalex.org/W3094502228","https://openalex.org/W3098649723","https://openalex.org/W3166021751","https://openalex.org/W3166398787","https://openalex.org/W4301312111"],"related_works":["https://openalex.org/W3194203264","https://openalex.org/W3212731309","https://openalex.org/W4385571814","https://openalex.org/W4226261765","https://openalex.org/W4417438331","https://openalex.org/W2994673210","https://openalex.org/W7167855795","https://openalex.org/W3123615524","https://openalex.org/W3194016324","https://openalex.org/W4414448144"],"abstract_inverted_index":{"Transformer":[0,25,47,108],"is":[1,44,102],"a":[2],"powerful":[3],"model":[4,48,68],"for":[5],"text":[6,118],"understanding.":[7],"However,":[8],"it":[9],"is\ninefficient":[10],"due":[11],"to":[12,16,67],"its":[13],"quadratic":[14],"complexity":[15],"input":[17],"sequence":[18],"length.":[19],"Although\nthere":[20],"are":[21,28],"many":[22],"methods":[23],"on":[24,31,50,78,96],"acceleration,":[26],"they":[27],"still":[29],"either\ninefficient":[30],"long":[32,117],"sequences":[33],"or":[34,115],"not":[35],"effective":[36,89],"enough.":[37],"In":[38,52,84],"this":[39,85],"paper,":[40],"we\npropose":[41],"Fastformer,":[42,53],"which":[43],"an":[45],"efficient":[46,105],"based":[49,77],"additive\nattention.":[51],"instead":[54],"of":[55],"modeling":[56,91,119],"the":[57],"pair-wise":[58],"interactions\nbetween":[59],"tokens,":[60],"we":[61],"first":[62],"use":[63],"additive":[64],"attention":[65],"mechanism":[66],"global\ncontexts,":[69],"and":[70,110],"then":[71],"further":[72],"transform":[73],"each":[74],"token":[75],"representation":[76],"its\ninteraction":[79],"with":[80,92],"global":[81],"context":[82,90],"representations.":[83],"way,":[86],"Fastformer":[87,101],"can\nachieve":[88],"linear":[93],"complexity.":[94],"Extensive\nexperiments":[95],"five":[97],"datasets":[98],"show":[99],"that":[100],"much":[103],"more":[104],"than\nmany":[106],"existing":[107],"models":[109],"can":[111],"meanwhile":[112],"achieve":[113],"comparable":[114],"even\nbetter":[116],"performance.":[120]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":1}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2021-08-30T00:00:00"}